Genetic Architecture of Behavioral Syndromes

With collaborators Ann Hedrick and Ned Dochtermann (NDSU), I am investigating how behavioral correlations arise and are maintained in Western Field crickets. We are in the process of comparing the strength of behavioral syndromes across 4 populations of crickets sampled throughout California, New Mexico and Arizona. We’ve just wrapped up the first part of the project on comparison of G matrices among our 4 populations (see our newest paper here). We are now using a large-scale artificial selection experiment to test whether behavioral syndromes respond to selection and are reversible.

Variation in Nest Architecture in Solitary Bees

With fellow postdoctoral researcher Bryan Helm (NDSU) and Rachel Malinger (USDA), I investigate how components of nesting behavior and nest architecture interact in the alfalfa leafcutter bee. We have found evidence for individual differences in nest building and that females show important variation in the amount of food provision they provide their offspring with (See our recent article here). We are also investigating how insecticide exposure affects nest building in this important crop pollinator. You can find more information on this project on our Crowdfunding page.

Statistical Approaches for Comparing Variance Components Among Datasets

Many studies rely on the estimation of variance components (among/within-individual variance, additive genetic variance) or variance ratios (repeatability, heritability, individual specialization) to make inference on rates of evolutionary changes or various ecological processes. In many cases, statistical methods allowing to directly compare changes in variance ratio or variance components are poorly defined. With Ned Dochtermann, we used simulation data to compare the power to detect changes in variance components and variance ratios across different statistical approaches (Bayesian and Frequentist). With this project, we hope to provide evolutionary ecologists with robust tools and guidelines allowing them to compare variation at multiple levels between datasets, populations, treatments or sexes. Our manuscript is now available as a preprint.